Matsusue Eiji, Inoue Chie, Shimoda Manabu, Nakamura Tomoya, Matsumoto Shota, Matsumoto Kensuke, Tanino Tomohiko, Nakamura Kazuhiko, Fujii Shinya
Department of Radiology, Tottori Prefectural Central Hospital, Tottori, Japan.
Department of Neurology, Tottori Prefectural Central Hospital, Tottori, Japan.
Acta Radiol. 2024 Jul;65(7):825-834. doi: 10.1177/02841851241253775. Epub 2024 May 24.
Brain magnetic resonance imaging voxel-based morphometry (VBM) and perfusion single-photon emission computed tomography (SPECT) are useful for differentiating dementia with Lewy bodies (DLB) from Alzheimer's disease (AD).
To determine whether combining multiple parameters of VBM and SPECT using a multiparametric scoring system (MSS) improves diagnostic accuracy in differentiating DLB from AD.
In total, 23 patients with DLB and 57 patients with AD underwent imaging using a voxel-based specific regional analysis system for AD (VSRAD), an easy Z-score imaging system, and a Z-Graph using three-dimensional stereotactic surface projection. The cutoff values were determined using the receiver operating characteristic curve to differentiate DLB from AD for all parameters. Patients were scored 1 (DLB) or 0 (AD) for each statistically significant parameter, according to a threshold. The total score was determined for each case to obtain a cutoff value for the MSS.
The mean Z-scores in the medial temporal lobes using the VSRAD were significantly lower in patients with DLB than in those with AD. Each Z-score of the summed Z-scores in all four segmented regions of the occipital lobes using the Z-Graph was significantly higher in patients with DLB than in those with AD. Among the five parameters, the highest accuracy was 80% for the Z-score of the summed Z-scores in the left medial occipital lobe. For the MSS, a cutoff value of four improved the diagnostic accuracy to 82%.
MSS was more accurate than any single parameter of VBM or SPECT in differentiating DLB from AD.
脑磁共振成像基于体素的形态测量(VBM)和灌注单光子发射计算机断层扫描(SPECT)有助于区分路易体痴呆(DLB)和阿尔茨海默病(AD)。
确定使用多参数评分系统(MSS)结合VBM和SPECT的多个参数是否能提高区分DLB和AD的诊断准确性。
总共23例DLB患者和57例AD患者使用针对AD的基于体素的特定区域分析系统(VSRAD)、简易Z评分成像系统以及使用三维立体定向表面投影的Z图进行成像。使用受试者工作特征曲线确定所有参数区分DLB和AD的临界值。根据阈值,对每个具有统计学意义的参数,患者被评为1分(DLB)或0分(AD)。确定每个病例的总分以获得MSS的临界值。
使用VSRAD时,DLB患者内侧颞叶的平均Z评分显著低于AD患者。使用Z图时,DLB患者枕叶所有四个分割区域的总Z评分的每个Z评分均显著高于AD患者。在这五个参数中,左内侧枕叶总Z评分的Z评分最高准确率为80%。对于MSS,临界值为4时将诊断准确性提高到82%。
在区分DLB和AD方面,MSS比VBM或SPECT的任何单个参数都更准确。